Exceptionally high trans (anti) stereoselectivity in catalytic

Daniel T. Nowlan III, Timothy M. Gregg, Huw M. L. Davies, and Daniel A. Singleton. Journal of the .... Huw M. L. Davies and Robert J. Townsend. The Jo...
1 downloads 0 Views 1MB Size
1906

J . Am. Chem. SOC.1990,112, 1906-1912

Exceptionally High Trans (Anti) Stereoselectivity in Catalytic Cyclopropanation Reactions Michael P. Doyle,* Vahid Bagheri, Thomas J. Wandless, Nancy K. Harn, David A. Brinker, Cassandra T. Eagle, and Kuo-Liang Loh Contribution from the Department of Chemistry, Trinity University, San Antonio, Texas 78284. Received July 6, 1989

Abstract: Exceptionally high trans (anti) stereoselectivitiesare obtained in rhodium(I1) carboxylate and carboxamide catalyzed alkene cyclopropanation reactions with 2,6-di-rert-butyl-4-methylphenyldiazoacetate (BDA). With monosubstituted ethylenes, use of rhodium(I1) acetamide provides a 10-2056 increase in the relative yield of the transdisubstituted cyclopropane. Compared to results from rhodium(I1) acetate catalyzed reactions with ethyl diazoacetate, the selectivity enhancement achieved with BDA corresponds to energy differences of between 1.0 and 2.0 kcal/mol and, in many cases, provides greater than 95% relative yield of the trans (anti) isomer. Intermediate selectivities are obtained with 2,3,4-trimethyl-3-pentyldiazoacetate (ODA) that are normally resistant and 3-isopropyl-2-methyl-3-heptyldiazoacetate (UDA). Alkenes such as 2,5-dimethyl-2,4-hexadiene to changes in their trans/cis cyclopropane product ratio by either catalyst modification or the use of alkyl diazoacetates such as ODA or UDA undergo highly selective cyclopropanation with BDA. Relative reactivities for cyclopropanation of selected alkenes extend over 3 orders of magnitude for reactions with BDA, and they suggest the extent of regioselectivity control in reactions with dienes and polyenes. Catalyst influences on reactivity and stereoselectivity clearly show the advantages of dirhodium(I1) tetrakis(acetamide) over rhodium(I1) carboxylates for the production of the trans (anti) cyclopropane isomer and for highly regioselective cyclopropanation reactions with BDA.

Among all of the catalysts that have been developed for carbene addition to multiple bonds, rhodium(I1) carboxylates are the most effective for intermolecular reactions that employ diazo carbonyl compounds.' Alkenes ranging from enol ethers and acetates through monoalkenes to dienes and trienes,2" but not a,D-unsaturated carbonyl compounds or nitriles,' undergo Rh2(OA~)~-catalyzed cyclopropanation (eq 1) in high yield with

minimum interference from normally competitive carbenoid dimerization.* However, stereoselectivitiesfor these transformations are low. In reactions with monosubstituted olefins, trans/cis cyclopropane ratios rarely exceed 2,and even cis-disubstituted alkenes exhibit anti/syn isomer ratios that are generally less than 4.3-5

We have recently reported9 that enhancement of trans (anti) stereoselectivities in cyclopropanation reactions could be achieved with the use of rhodium(I1) acetamide'O as the catalyst and 2,3,4-trimethyl-3-pentyl diazoacetate (ODA) as the carbenoid precursor. The electronic influence of the acetamide substituents in Rh,(acam), apparently increases the stability of the intermediate metal carbene, thereby effecting a closer approach of the unsaturated substrate to the electrophilic carbenoid center." The larger steric bulk of the diazo ester favors the orientation of reactants that leads to the trans (anti) isomer. Still, the combined influence of these modifications is limited. They increase the ( I ) (a) Doyle, M. P. Chem. Reu. 1986,86,919. (b) Maas, G. Top. Curr. Chem. 1987, 137,75. ( 2 ) (a) Hubert, A. J.; Noels, A. F.; Teyssie, P. Synthesis 1976, 600. (b) Anciaux, A. J.; Hubert, A. J.; Noels, A. F.; Petinoit, N.; Teyssie, P. J. Org. Chem. 1980,45, 695. (3) Doyle, M. P.; Dorow, R. L.; Buhro, W.E.; Griffin, J. H.; Tamblyn, W.H.; Trudcll, M. L. Organometallics 1984, 3, 44. (4) Doyle, M. P. Acc. Chem. Res. 1986, 19, 348. (5) Piers, E.;Moss, N. Tetrahedron Lett. 1985, 26, 2735. (6) Doyle, M. P.; Dorow, R. L.; Tamblyn, W.H.; Buhro, W. E. Tetrahedron Lett. 1982, 23, 2261. (7) Doyle, M. P.; Dorow, R. L.; Tamblyn, W . H. J . Org. Chem. 1982, 47, 4059. (8) Doyle, M. P.; van Leusen, D.; Tamblyn, W . H. Synthesis 1981, 787. (9) Doyle, M. P.; Loh, K.-L.; DeVries, K.M.; Chinn, M. S.Tetrahedron Lett. 1987, 28, 833. (10) (a) Ahsan, M. Q.;Bernal, I.; Bear, J. L. Inorg. Chem. 1986,25,260. (b) Best, S.P.; Chandley, P.; Clark, R. J. H.; McCarthy, S.;Hunthouse, M. B.; Bates, P. A. J. Chem. Soc., Dalton Trans. 1989, 581. (11) Doyle, M. P.; Griffin, J. H.; Bagheri, V.; Dorow, R. L. Organometallics 1984, 3, 53.

relative percentages of the thermodynamically more stable cyclopropane isomer from 62% to 87% (with styrene) and from 79% to 92% (with cyclohexene)? which corresponds to only a 0.6-0.8 kcal/mol energy difference from stereoselectivities achieved with Rh2(OAc), and ethyl diazoacetate (EDA). We now wish to report the optimization of this methodology that allows catalytic cyclopropanation to be achieved with trans (anti) stereoselectivities that are greater than 90% and, in many cases, exceed 95%.

Results Influence of Diazo Esters. On the basis of the enhancement in relative yields of trans (anti) cyclopropane stereoisomers obtained with 2,3,4-trimethyl-3-pentyl diazoacetate (ODA)?J2 two additional diazoacetates with bulky ester appendages, 3-isopropyl-2-methyl-3-heptyldiazoacetate (UDA)12 and 2,6-di-tertbutyl-4methylphenyl diazoacetate (BDA), were prepared in order to determine their influence on stereoselectivity. Both ODA and

UDA

ODA

UDA were conveniently synthesized from the corresponding alcohols in three steps by diketene addition,I3*I4diazo transfer,'$ and deacylation,'6 and ODA was sufficiently stable to be purified by low-temperature distillation. The synthesis of BDA was accomplished in two steps by performing diketene addition to 2,6di-tert-butyl-4-methylphenol (BHT) in the presence of methanesulfonyl azide (eq 2) with significant advantages in product NaOAc

Nz

t.Bu

L

yield and purity over the sequential three-step process. The relatively high melting point of BDA (1 53 "C) suggests its exceptional stability. (12) The 8-carbon alkyl attachment is denoted octyl in ODA; the 11carbon alkyl attachment is undecyl in UDA. 113) Clemens. R. J. Chem. Reu. 1986.86. 241. (14) Doyle, hi.P.; Dorow, R. L.; Terpstra, J. W.;Rodenhouse, R. A. J. Org. Chem. 1985, 50, 1663. (15) Taber, D. F.; Ruckle, R. E., Jr.; Hennessy, M. J. J. Org. Chem. 1986, 51, 4077. (16) Doyle, M. P.; Shanklin, M.S.;Pho, H. Q.; Mahapatro, S. N. J. Org. Chem. 1988, 53, 1017.

0002-7863/90/15121906$02.50/0 0 1990 American Chemical Society

High Trans (Anti) Stereoselectivity in Cyclopropanation Table I. Stereoselectivities in Rh,(OAc),-Catalyzed Cyclopropanation Reactions with ReDrescntative Alkenes’ - % trans (anti)* EDA BDA alkene (1)‘ (4) 62 67 (92) 70 (88) 71 (96) ethyl vinyl ether (a) 60 71 (90) 74 (87) 75 (93) 1-hexene (b) 62 71 (94) 73 (87) 84 (94) styrene (c) 75 84 (89) 89 (84) 93 (92) 3,3-dimethyl-I-butene (a) 97 (79) 69 (52)/ 72 cyclopentene (e) 79 90 (86) 92 (81) 91 (60p cyclohexene (f) 67 76 (9oj 82 (84j 95 (92) bicyclo[2.2.l]hept-2-ene (9) 87 85 (88) 97 (92) dihydropyran (b) 60 64 (86) 65 (81) 96 (89) 2-methyl-Zbutene ( i ) 2.5-dimethvl-2.4-hexadiene 6) 62 65 (88) 60 (73) 94 (88) “Reactions were performed at room temperature with 1.0 mol % of the catalyst, based on the diazo compound, and a 10-fold molar excess of the alkene. Yield of cyclopropane products rior to chromatographic purification in parentheses. CDatafrom ref 3. Stereoselectivities are consistent with those reported in ref 9. Only trace amounts of lactone 7 were observed. ‘Lactones 8 and 9 were formed competitively (5-18% yield). /Complex mixture consisting of cyclopropane derivatives (34%) and insertion products (66%). ‘Complex mixture consisting of cyclopropane derivatives (14%) and insertion products (86%).

B

Cyclopropanation reactions were performed by controlled addition of the diazo ester to the combination of alkene and 1.0 mol % Rh2(OAc)4in dichloromethane at room temperature. A 10-fold excess of the alkene was employed to obtain uniformity in yield determinations by minimizing competitive carbene dimer formation. However, comparable product yields could be obtained in reactions with the most reactive alkenes by using only a 2-fold excess of alkene and adjusting the rate of addition. Stereoselectivities in cyclopropane formation from Rh2(OAc)4-catalyzed reactions of EDA, ODA, UDA, and BDA with a representative series of mono-, di-, and trisubstituted ethylenes are reported in Table I. Stereoisomers were identified by their characteristic NMR spectra, from which coupling constants, especially for the hydrogen on carbon adjacent to the carboxylate group, revealed stere~chemistry.~~~ As expected, the change in cyclopropanation selectivities between EDA and ODA is greater than that between ODA and UDA. Compared to the use of tert-butyl diazoacetate,” use of ODA has a greater influence on sterwlectivity, and even though far removed from the reaction center, the n-butyl group of UDA has a measurable influence on the relative percentage of the trans (anti) isomer compared to the methyl group of ODA. Perhaps the most surprising result, however, is the relative absence of stereoselectivity enhancement in reactions of either ODA or UDA with the trisubstituted ethylenes 2-methyl-2-butene and 2,S-dimethyl-2,4-hexadiene. The phenolic ester BDA, whose development was influenced by the enhanced stereoselection in aldol condensation reactions observed by Heathcock and co-workers with BHT esters,Is is superior to either ODA or UDA for stereoselectivity enhancement in cyclopropanation reactions. Relative yields of the trans (anti) stereoisomers (4) are the highest yet achieved in catalytic reactions, and even the normally resistant trisubstituted ethylenes are very responsive to the steric bulk of the phenolic attachment. Indeed, the exceptional selectivity obtained in the preparation of 4j makes this the method of choice for the synthesis of trans chrysanthemic acid derivative^.'^ However, cyclohexene gave abnormally low yields of cyclopropane products with BDA due to competitive formation of its apparent allylic insertion product 5. Similarly, 6 was formed in reactions with cyclopentene, but the analogous product from reactions with dihydropyran was formed in low yield ( styrene (47%) > cyclohexene (21%). In addition, with N-alkyl substituents larger than methyl, the carbenoid reagents are subject to intramolecular C-H insertion.21 In Rh2(OAc)4-catalyzed reactions of styrene with N,N-diisopropyldiazoacetamide,for example, the trans-2phenylcyclopropanecarboxamide isomer was obtained in 98% relative yield, but the actual yield of cyclopropane products could not be made to exceed 53% because of competitive intramolecular C-H insertion (eq 4). Reactions of N,N-diisopropyldiazoacet-

amide in the presence of cyclohexene yielded only the intramolecular C-H insertion products. Similar insertion processes were unimportant for Rh,(OA~)~-catalyzed reactions with ODA and BDA performed in the presence of alkenes. Influence of the Catalyst. A variety of dirhodium(I1) compounds with differing electronic influences imparted on the rho(20) Dave, V.; Warnhoff, E . Org. Read. (N.Y.) 1970, 18, 217. (21) Doyle, M. P.; ShanWin, M. S.;Oon, S.-M.; Pho, H. Q.;Van der Heide, F. R.; Veal, W. R.J. Org. Chem. 1988, 53, 3384.

Doyle et al.

1908 J. Am. Chem. Soc., Vol. 1 1 2, No. 5, 1990 Diazoacetate Cyclopropanation Stereoselectivities' % trans (anti)b catalvst ethvl vinvl ether stvrene cvclohexene Rhi(NHCOCH3)d 72 68 89 Rh2(00CCH,)d 62 62 79 RhAOOCCiH,), 60 60 77 Rh;(NHCOCF;), 62 59 74 Rh2(OCSCH,)d 60 62 75 Rh2(00CCF,), 59 56 67 Rh2(00CC,F,)f 57 52 66 a Reactions were performed at room temperature with 1 .O mol % of the catalyst. based on the diazo compound, and a IO-fold molar excess of the alkene. bProduct yields ranged from 65% to 98% with median yields of 8248%. cPercent trans isomer from reactions with 3,3-dimethyl-I-butene (60%), bicyclo[2.2.l]hept-2-ene (59%), I-hexene (55%), and 2,5-dimethyl-2,4-hexadiene(58%). ,

I

dium(I1) center by its ligands were prepared to develop one that would exhibit, relative to Rh2(OAc),, increased carbenoid stereoselectivity in cyclopropanation reactions. Selectivity evaluations were made from reactions of EDA with a selection of olefins that provide a range of reactivities toward carbenoid addition, and the results of this survey are presented in Table 11. As is evident from these results, rhodium(I1) acetamide (acam) provides the highest and rhodium(I1) perfluorobutyrate (pfb) the lowest trans (anti) cyclopropanation selectivities. Rhodium(I1) acetamide has previously been prepared as a mixture of the tetraacetamide with varying amounts of Rh2( a ~ a m ) ~ ( O Aand c ) R h , ( a ~ a m ) ~ ( O A cby ) ~treating Rh2(OAc), in a melt of acetamide.'& Even when acetamide was employed in 100-150-fold molar excess over Rhz(OAc), and the procedure repeated several times, pure Rh2(acam)4could be obtained only by HPLC separation. We have repeated these experiments, with Rh2(OAc), and with the more labile Rh2(00CCF3), as reactants,u8 and obtained similar results. However, when this synthesis was performed in refluxing chlorobenzene under conditions where acetic acid was trapped by sodium carbonate in a Soxhlet extraction apparatus, pure Rh2(acam), was obtained without any evident trace of Rh,(acam),(OAc). Spectral analysis confirmed that the tetraacetamide compound had the cis g e ~ m e t r y . ' ~ . ~ ~ Stereoselectivities in cyclopropane formation from Rh2(scam),-catalyzed reactions of EDA, ODA, UDA, and BDA with a representative series of alkenes are reported in Table 111. Even with EDA, Rh2(acam), increases the relative yield of the trans (anti) isomer, compared to that obtained with R ~ , ( O A C ) by ~ , as much as 15%, and the use of ODA provides additional enhancement. Further improvements are seen with UDA, but it is BDA from which the highest trans (anti) stereoselectivities are achieved in catalytic cyclopropanation reactions. With Rh2(acam), as the catalyst, products from competitive intramolecular carbon-hydrogen insertion of ODA and UDA are observed, especially in reactions performed with 1-hexene, cyclohexene, and 2,5-dimethyl-2,4-hexadiene.With ODA as the reactant, lactone 7 is formed, whereas with UDA both 8 and 9 are produced (8/9 = 2.0). Their yield is dependent on the alkene

4

H,CH,(

1

8

Stereoselectivities in Rh2(acac),-CatalyzedCyclopropanation Reactions with Representative Alkenes' % trans (anti)b EDA ODA UDA BDA alkene (1)c (3)' (4) 85 (96) ethyl vinyl ether (a) 72 80 (86) 63 1-hexene (b) 81 (84) 83 (55) 93 (88) styrene (c) 68 87 (81) 93 (78) 98 (90) 3.3-dimethyl-I-butene (d) 83 89 (80) 99 (90) cyclohexene (f) 89 92 (75) 96 (39) 92 (21)' bicyclo[2.2.l]hept-2-ene (9) 82 83 (86) 86 (63) 92 (78) 2-methvl-2-butene (il 58 66 172) 65 (671 2,5-dimethyl-2,4-hexadiene(j) 60 59 (75) 60 (62) 98 (75) Reactions were performed at room temperature with 1 .O mol % of the catalyst, based on the diazo compound, and a IO-fold molar excess of the alkene. bProduct yield in parentheses. CDatafrom ref 9. dLactone 7 was produced in 8-22% yield. CLactones8 and 9 were formed in 15-50% yield. /Complex mixture consisting of cyclopropane derivatives (18%) and insertion products (82%). Table 111.

Table 11. Influence of Dirhodium(I1) Catalysts on Ethyl

A

CY

9

employed, but increasing the concentration of the alkene lessens the relative importance of this competing process. The scope of this intramolecular process is currently under investigation and will be described in a subsequent publication. (22) (a) Doyle, M. P.; Colsman, M. R.; Chinn, M. S.Inorg. Chem. 3984, 23, 3684. (b) Doyle, M. P.; Mahapatro, S.N.; Caughey, A. C.; Chinn, M. S.;Colsman, M. R.; Harn, N. K.; Redwine, A. E. Ibid. 1987, 26, 3070. (23) Ahsan, M. Q.;Bernal, I.; Bear, J. L. Inorg. Chem. Acra 1986, 115, 135.

(ad

Relative Reactivities. To determine if the increases in stereoselectivity occur with changes in reactivity, experiments were performed with pairs of alkenes, normally including 1-hexene, to obtain relative reactivities for interception of the intermediate metal carbene, and these results are reported in Table IV. Prior determinations that were limited to cyclopropanation reactions with EDA catalyzed by Rh2(0Ac),*' showed that aliphatic monosubstituted alkenes were less reactive than di- or trisubstituted alkenes and that vinyl ethers were more reactive than styrene, which, in turn, was more reactive than 1-hexene. This same reactivity scale, although moderately expanded, is observed in reactions of EDA catalyzed by Rh2(acam),. With either ODA or BDA, however, 2,5-dimethyl-2,4-hexadieneis no longer more reactive than 1-hexene, and cyclohexene is considerably less reactive than 1-hexene in cyclopropanation reactions performed with BDA. Cross-checks of relative reactivity values with different pairs of alkenes were performed to certify the validity of these numbers. These relative reactivities were used to evaluate the potential of BDA for regioselective cyclopropanation of 4-vinylcyclohexene (eq 5 ) . On the basis of the data in Table IV for Rh2(OAc),+ BDA

Rh2L4

t-Bu

11

t-BL

10

catalyzed reactions of BDA, the expected distribution of products would be 93% 10 and 7% 11; the actual distribution was 93% 10 (83% trans, 17% cis) and 7% 11. Similarly, with Rh2(acam), the product distribution from the reaction of BDA with 4-vinylcyclohexene was 98% 10 and 2% 11, with a stereoisomer ratio for 10 of 67 (trans/&).

Discussion Stereoselectivities obtained by increasing the steric bulk of diazoacetate esters are consistent with the mechanism for cyclopropanation that is outlined in Scheme I. As previously reported,lsv4J1 stabilization of the transition states leading to cyclopropane products by association of the ester carbonyl oxygen with the developing electropositive center of the reacting alkene accounts for the preferential trans (anti) stereoselectivity that is observed in intermolecular cyclopropanation reactions of diazo esters, as well as for dihydrofuran formation in reactions of vinyl ethers with diazo and for the apparent allylic carbon-hydrogen insertion of diazomalonate esters with alkene^.'^.^^ (24) (a) Wenkert, E.; Alonso, M. E.; Buckwalter, B. L.; Sanchez, E. L. J . Am. Chem. Soc. 1983, 105, 2021. (b) Alonso, M. E.;Morales, A.; Chitty, A. W. J. Org. Chem. 1982,47,3748. (c) Alonso, M. E.; Jano, P.; Hernandez, M. I.; Greenberg, R. S.;Wenkert, E. Ibid. 1983, 48,3047.

J. Am. Chem. SOC.,Vol. 112, No. 5, 1990 1909

High Trans (Anti) Stereoselectivity in Cyclopropanation

Table IV. Relative Reactivities in Rh,(OAc),- and Rh,(acam),-Catalyzed Cyclopropanation Reactions of Representative Alkenes with EDA,

ODA, and BDA EDA alkene

Rhz(OAc),

RhZ(acam),

8.3 3.5 1.o 2.1 2.5

156

ethyl vinyl ether

relative reactivities' ODA RhZ(OAc), Rhz(acam), 9.0 1.9 1.o 0.26 1.2

BDA Rh2(OAc),

Rh2(acam),

13 3.8 1.o 0.28 0.07F

19b

206

14 IO 3.9 styrene 1.o 1.o 1 .o 1 -hexene 0.29 0.090 2,5-dimethyl-2,4-hexadiene 2.0 0.022c 1.o 0.73 cvclohexene 4Reactions performed in dichloromethaneat 25 OC [with Rh2(OAc),] or at reflux [with Rh,(acam),] with a IO-fold molar excess of each alkene and 1.0 mol W of the catalyst. bBecause of its volatility, ethyl vinyl ether was replaced by n-butyl vinyl ether in these experiments. CCalculation includes cyclopropane products and the insertion product 5.

Table V. Relative Reactivities for Trans (Anti) and Cis (Syn) Cyclopropanationwith EDA, ODA, and BDA Catalyzed by Rh,(OAc), and

Rh2(acam), EDA Rh,(OAc), trans cis (anti) (syn)

alkene 8.6 ethyl vinyl ether 3.6 styrene 1.0 I -hexene 2,5-dimethyl-2,4-hexadiene 2.2 cvclohexeneb -, ------ -----

3.3 .~

5.3 2.2 0.67 1.3 0.83 ~

~~

Rh2(acam), trans cis (anti) (syn) 17 11 1.0 1.9 1.4

6.7 5.1 0.59 1.3 0.19

relative reactivities' ODA Rh,(OAc), Rh,(acam), cis cis trans trans (anti) (syn) (anti) (syn) 8.5 1.9 1.0 0.24 1.5

4.2 0.78 0.41 0.13 0.17

19 4.2 1.0 0.21 0.83

4.6 0.62 0.23 0.15 0.072

BDA Rh2(0Ac)4

trans (anti) 12 4.3 1.0 0.35 0.013

cis (syn) 5 .O 0.81 0.33 0.022 0.0012

Rh,(acam), trans cis (anti) (syn) 18 15 1.o 0.095 0.0039

3.2 0.30 0.075 0.0019

0.0003

OData calculated from Table IV. The relative reactivities for trans cyclopropane formation from I-hexene were set at 1.0, and relative reactivities for cis cyclopropane formation for 1-hexene were calculated from the observed stereoselectivities. Relative reactivities for cyclopropanation of alkenes were determined from the formula [relative reactivity of alkene (from Table IV)] X [sum of relative reactivities for cyclopropanation of I-hexene (Table V)] X [mole fraction of stereoisomer from cyclopropanation of alkene (from Tables I and Ill)]. bValues for reactions with BDA are dependent on the relative yield of 5, and relative reactivities for apparent insertion are approximately 0.081 with Rhz(OAc), and 0.016 with Rhz(acam),. Scheme I

Increasing the size of Z (Scheme I) favors the transition state leading to the trans cyclopropane isomer. This effect is more pronounced in reactions with less electron rich alkenes (compare 1-hexene with ethyl vinyl ether), presumably because they require a closer approach to the carbenoid center in the transition state for cyclopropanation. The same can be said for the influence of the catalyst; use of rhodium(I1) acetamide provides greater selectivity in intermolecular cyclopropanation reactions because its derivative metal carbene complexes are more stable than those derived from rhodium(I1) acetate, which, in turn, are more stable than those formed from rhodium( 11) perfluorobutyrate. Increases in trans (anti) selectivity for cyclopropanation of alkenes are a function of their substituents. Cis-disubstituted alkenes such as cyclohexene are most responsive to changes in the steric bulk of the diazo ester, and trisubstituted alkenes such as 2,5-dimethyl-2,4-hexadieneare least responsive. However, electronic influences governing closeness of approach of the alkene to the carbene center, influenced both by alkene substituents and P.; Hernandez, M. I. J . Org. Chem. 1980, S.;McDaniel, R.S.,Jr.; Peace., B. W. Terrahe-

(25) (a) Alonso, M. E.; Jano,

45, 5299. (b) Wulfman, D. dron 1976, 32, 124 1.

by ligands of the catalyst, are of comparable importance. Calculation of the relative rates for cyclopropanation yielding trans (anti) and cis (syn) isomers, each based upon the formation of the trans cyclopropane product from reactions with 1-hexene (Table V), provides a guide to these influences. Relative reactivities for formation of trans (anti) cyclopropane isomers are of comparable magnitude for reactions with EDA, and their breadth of values increases by only a small degree in reactions performed with ODA. With BDA, however, relative reactivities for trans (anti) cyclopropanation of the selected series of alkenes vary by over 3 orders of magnitude, presumably because of steric constraints placed upon alkene orientation by the reactive carbene in the transition state. Rhodium(I1) acetamide expands the range of values, relative to rhodium(I1) acetate, in each series by a factor of 2 or more. Among monosubstituted alkenes, however, reactivities are relatively invariant with respect to the size of the diazo compound. Relative reactivities for formation of the cis (syn) isomer, on the other hand, generally exhibit substantial changes with the increased size of the diazo ester, and these changes are most pronounced with Rh,(acam), as the catalyst. Nowhere are these influences more evident than in reactions with 2,5-dimethyl-2,4hexadiene, where a 10-fold decrease in reactivity is observed between EDA and ODA and another 80-fold decrease occurs between ODA and BDA. Even the reactivity of styrene for formation of its cis-disubstituted cyclopropane derivative decreases from 5.1 to 0.30 relative to 1-hexene trans reactivity. The formation of allylic insertion products is the dominant pathway in reactions of cyclohexene with BDA. Reactivities for insertion relative to those for cyclopropanation are 2.7 with Rh2(OAc), and 2.8 with Rh,(acam)+ A similar process has been observed in copper-catalyzed reactions of cyclohexene with diazomalonate esters,% but its contribution to total product formation was minor. A few additional examples have been reported,''*24a but none in yields as high as those encountered from reactions (26) Wulfman, D. S.;McGibboney, B. G.; Steffen, E. K.;Thinh, N. V.; McDaniel, R. S.Jr.; Peace, B. W. Tetrahedron 1976, 32, 1257.

1910 J . Am. Chem. Soc., Vol. 112, No. 5, 1990

of BDA with cyclohexene. This process has the appearance of an intermolecular addition-elimination reaction,la but further investigations are required to confirm t h e mechanism. Not only do the composite results document effective procedures t o achieve high trans (anti) selectivities in intermolecular alkene cyclopropanation reactions but also relative reactivities indicate t h e extent t o which regioselective cyclopropanation can be achieved. With 4-vinylcyclohexene, regioselectivities with the use of BDA can be predicted from Table IV, and additional examples are anticipated t o show similar correspondence. Cyclopropane products derived from reactions with BDA a r e amenable to functionalization a t the position a to the carbonyl group and can be employed for the synthesis of a-amino acids.27 Experimental Section General Methods, NMR spectra were obtained on a Varian VXR-300 spectrometer; chemical shifts are reported in 6 units with tetramethylsilane as the internal standard. Infrared spectra were recorded on an IBM IR/32 FT spectrometer, and mass spectra were obtained with the Hewlett-Packard 5995C GC/MS system operated at 70 eV. Analytical gas chromatographic analyses were performed on a Hewlett-Packard 5890A capillary GC with use of either SP-2330 or methylsilicone columns or both. Elemental analyses were performed by Texas Analytical Laboratories, Inc. Rhodium(I1) acetate was prepared from rhodium trichloride;28 rhodium(I1) butyrate, trifluoroacetate, and perfluorobutyrate were synthesized by acetate displacement from stock rhodium(11) acetate in the refluxing carboxylic acid that contained the corresponding anhydride.s Rhodium(I1) trifluoroacetamide was obtained by heating rhodium(l1) acetate in a melt of trifluoroacetamide at 140-150 OC according to the orocedure of Bear.N Rhodium(I1) thioacetate was prepared and characterized according to the literature procedure." Alkenes were commercially available and were purified by distillation prior to use. Methanesulfonyl azide was prepared from methanesulfonyl chloride and sodium azide.', 2,3,4-Trimetbyl-3-pentyl Diazoacetate (ODA). A mixture of 2,3,4trimethyl-3-pentanol (9.75 g, 75.5 mmol), prepared by methyllithium addition to 2,4-dimethyl-3-pentanone,and anhydrous sodium acetate (0.043 g, 0.52 mmol) was heated to 90 OC in a 100-mL round-bottom flask fitted with an addition funnel containing diketene (6.98 g, 83.1 mmol). Addition of diketene to the rapidly stirred alcohol was allowed to occur dropwise over 20 min, and the resulting mixture was maintained at 90 OC for an additional 1 h. Distillation of the resulting acetoacetic ester (bp 42-44 OC at 0.05 Torr) yielded 9.45 g of a colorless liquid (44.7 mmol,58% yield). 'H NMR (CDCI,, 300 MHz): 6 4.93 (s, enol CH, 5%), 3.38 (s, keto CHI, 95%), 2.67 (s, enol CH,), 2.27 (s, keto CHI), 2.26 (hept, J = 7.0 Hz, 2 H), 1.43 (s, ester CH, of keto form), 1.41 (s, ester CH, of enol form), 0.95 (d, J = 7.0 Hz, 6 H of keto form), 0.93 (d, J = 4.0 Hz, 6 H of keto form), and 0.92 and 0.88 (2 d, J = 7.0 Hz, 12 H of enol form). 2,3,4-Trimethyl-3-pentyl 3-oxobutanoate (6.40 g 30.3 mmol) was dissolved in 50 mL of anhydrous acetonitrile containing triethylamine (3.64 g, 36.0 mmol), and a solution of methanesulfonyl azide (4.48 g, 37.0 mmol) in 40 mL of acetonitrile was added dropwise to the stirred solution over 20 min. The resulting yellow solution of the diazoacetoacetate was maintained at room temperature for 4 h, and then 50 mL of 8% aqueous potassium hydroxide was added to effect acetyl cleavage. After being stirred for 24 h at room temperature, the solution was diluted by addition of water, the aqueous acetonitrile solution was washed with three 5 0 " portions of ether, and the combined ether solution was dried over anhydrous magnesium sulfate and concentrated under reduced pressure. ODA was purified by vacuum distillation (bp 84-87 "C at 0.1 Torr) to yield 3.46 g (17.4 mmol, 60% yield) of a brilliant yellow liquid. IH NMR (CDCI,, 300 MHz): 6 4.61 (s, CHN,), 2.28 (hept, J = 6.9 Hz, 2 H), 1.43 (s, 3 H), 0.96 (d, J = 6.9 Hz, 6 H), and 0.94 (d, J = 6.9 Hz, 6 H). "C NMR (CDCI,, 75 MHz): 8 160.5,94.2,92.0,46.5, 34.7, 28.1, 18.2, and 17.9. IR (neat): 2106 (C=N,) and 1694 ( C 4 ) cm-I. Anal. Calcd for CIOH18NZ02: C, 60.56; H, 9.17; N, 14.13. Found: C, 60.61; H, 9.24; N, 13.94. (27) Seebach, D.; Haner, R.; Vettiger, T. Hela Chim. Acta 1987,70, 1507. (28) Rampel, G. A.; Legzdins, P.; Smith, H.; Wilkinson, G. Inorg. Synrh. 1972, 13, 90. ( 2 9 ) Drago, R. S.;Long, J. R.; Cosmano, R. Inorg. Chem. 1982,21,2196. (30) Dennis, A. M.; Korp., J. D.; Bernal, I.; Howard, R. A.; Bear, J. L. Inorg. Chem. 1983, 22, 1522. (31) Baranovskii, I. B.; Bolubnichaya, M. A.; Mazo, G. Ya.; Nefedov, V. 1.; Salyb, Ya. V.; Shchelokov, R. N. Zh. Neorg. Khim. 1976.21, 1085; Russ. J. Inorg. Chem. (Engl. Transl.) 1976, 21, 591. (32) Boyer, J. H.; Mack, C. H.; Goebel, W.; Morgan, L. R.,Jr. J . Org. Chem. 1959, 24, 1051.

Doyle et al. 2-Methyl-3-isopropyl-3-heptylDiazoacetate (UDA).Diketene (9.28 g, I10 mmol) was added dropwise over 30 min to a rapidly stirred mix-

(1 7.23 g, 100 mmol), prepared ture of 2-methyl-3-isopropyl-3-heptanol by n-butyllithium addition to 2,4-dimethyl-3-pentanone,and sodium acetate (0.043 g, 0.5 mmol) that was maintained at 105 "C in a 250" round-bottom flask. After being stirred for an additional 90 min, the resulting acetoacetic ester was distilled (bp 91-94 OC at 0.1 Torr) to yield 15.0 g of a colorless liquid (58.6 mmol, 59% yield). IH NMR (CDCI,, 300 MHz): 6 4.95 (s, enol CH, 8%), 3.41 (s, keto CH,, 92%), 2.67 (s, enol CH,), 2.39 (hept, J = 7.0 Hz, 2 H), 2.28 (s, 3 H), 1.99-1.91 (m, 2 H), 1.32-1.24 (m, 4 H), 1.00 (d, J = 7.0 Hz, 6 H), 0.97 (d, J = 7.0 Hz, 6 H), and 0.90 (t, J = 7.1 Hz, 3 H). Diazo transfer from methanesulfonyl azide (9.68 g, 80.0 mmol) to 2-methyl-3-isopropyl-3-heptyl 3-oxobutanoate (17.92 g, 70.0 mmol) was effected with the use of triethylamine (8.08 g, 79.9 mmol) in anhydrous acetonitrile at 40 OC (12 h), and acetyl cleavage was performed by using 40 mL of 15% aqueous potassium hydroxide according to the procedure previously described for the preparation of ODA. After isolation, UDA was purified by column chromatography on neutral alumina with dichloromethane as the eluent to yield 13.52 g of a clear yellow liquid (56.7 mmol, 81% yield). IH NMR (CDCI,, 300 MHz): 8 4.62 (s, CHN,), 2.41 (hept, J = 7.0 Hz, 2 H), 2.00-1.92 (m,2 H), 1.36-1.24 (m, 4 H), 1.01 (d, J = 7.0 Hz, 6 H), 0.99 (d, J = 7.0 Hz, 6 H), and 0.91 (t, J = 7.0 Hz, 3 H). I3C NMR (CDCI,, 75 MHz): 6 160.2, 93.2, 46.4, 34.4, 33.6, 26.9, 23.6, 19.0, 18.9, and 14.1. IR (neat): 2106 (C=N2) and 1696 (C=O) cm-I. Anal. Calcd for Cl3HZ4N2O2:C, 64.96; H, 10.06; N, 11.66. Found: C, 64.87; H, 10.13; N, 11.60. S6-Di-tert-butyl-4-mtbylphenyl Diazoacetate (BDA).Diketene (4.20 g, 50.0 mmol) in 5.0 mL of anhydrous acetonitrile was added dropwise over 30 min to a rapidly stirred solution of 2,6-di-tert-butyl-4-methylphenol (5.51 g, 25.0 mmol),sodium acetate (0.21 g, 2.6 mmol), and methanesulfonyl azide (3.94 g, 32.6 mmol) in 20 mL of refluxing acetonitrile contained in a 100-mL three-neck flask fitted with a reflux condenser and an addition funnel. After addition was complete, the resulting brown solution was cooled to room temperature and stirring was continued for 7 h. The diazoacetoacetate product was isolated by adding water and extracting with ether, washing the ether extract with a 15% aqueous potassium hydroxide solution, and then drying the extract over anhydrous magnesium sulfate. Evaporation of the ether left a brown oil that was subjected to acetyl cleavage. Alternatively, addition of 30 mL of 15% aqueous potassium hydroxide solution precipitated a yellow powder that was filtered, washed with water, and dried. The resulting light yellow solid was identified as 2,6-di-tert-butyl-4-methylphenyldiazoacetoacetate (6.30 g, 19.0 mmol, 76% yield; mp 129-131 "C). 'H NMR (CDCI,, 300 MHz): 8 7.13 (s, 2 H), 2.51 (s, CH,CO), 2.33 (s, 3 H), and 1.34 (s, 18 H). A trace amount of BHT (